IBM’s AI-Based Personalized Pricing and Offers Boosts Malaysia Airlines’ Bookings

by 
David Kaplan
Thursday, October 3, 2019
 • 
7
 min read

Malaysia Airlines fliers who received personalized recommendations made over 20 percent more bookings, creating 44 percent more revenue, noted IBM’s Dee Waddell.

IBM and Malaysia Airlines have launched a partnership program designed to demonstrate that artificial intelligence is a practical tool for driving higher bookings and revenue through automation and targeted, dynamic pricing.

Even the name of IBM’s effort – it’s called “Personalized Pricing and Offers,” or PPO – is meant to evoke a plain-spoken and actionable product. This isn’t about experimentation for a distant future. It’s about now, Dee Waddell, IBM’s Global Managing Director, Travel & Transportation Industries, told Kambr Media.

The results, which were unveiled at last month’s World Aviation Festival, were based on IBM’s AI algorithms analyzing “hundreds of data points that could produce billions of possibilities to dynamically deliver” real-time offers in an email-based ad campaign.

Malaysia Airlines fliers who received the AI-generated email marketing program’s targeted recommended fares made 34 percent more bookings than customers in a control group who were not presented with the promotions.

The real-time booking recommendation effort also drove 54 percent more bookings in business class, Waddell said.

Ultimately, the test drove 44 percent greater revenue compared to the control group.

In addition to the numbers, the ability of AI to “free up” manual revenue management processes held the most appeal to Malaysia Airlines.

“Try to force your traditional marketers to take a step back and let the system start learning, and you’ll see quickly that it’ll outperform [previous methods], said Malaysia Airlines Digital Head Peter Pohlschmidt. “And don’t feel bad that it outperforms what you’ve been doing manually.”

In making the wider pitch of PPO’s value to airlines, Waddell said that effort it not limited to channels like email.

The capabilities can apply to every point of engagement –whether it’s done through self-service, airline staff, or via a vendor/partner.

We caught up with Waddell at the event and asked him to walk us through the role of AI in airline bookings and pricing, its open-source approach, as well as how the development of PPO also reflects IBM’s own wider and continual digital transformation.

Malaysia Airlines' Peter Pohlschmidt and IBM's Dee Waddell at the World Aviation Festival

Kambr Media: How does the partnership with Malaysia Airlines reflect the way airlines are integrating artificial intelligence into their revenue management and booking systems?

Dee Waddell:  AI is helping airlines make much better use of data, and airlines can differentiate themselves by applying AI to sift through large amounts of data to improve products, solve problems, personalize customer experiences and, overall, drive better business outcomes. We worked with Malaysia Airlines to help architect a more convenient and personalized experience for travelers shopping for airfares that would leverage IBM’s AI-based Personalized Pricing and Offers solution.

Malaysia Airlines did multiple tests with IBM to find out how to fill their seats with higher-level revenue buckets. We helped them put the data in a format, take it through the booking engine. We’ve started out with Malaysia in a production environment and we're taking this to scale. The outcome is more personalized destination recommendations for their customers, and, during the pilot phase, we’ve seen positive change in not only how customers book their airfare but also in the airline’s revenue. During the pilot, customers who received personalized recommendations made over 20 percent more bookings, creating 44 percent more revenue.

How has IBM itself evolved to better provide these kinds of solutions?

IBM, as part of our own transformation, has shifted its focus into a few key themes. One is, of course, we go to market by industry. We have 12 industries that we support. Travel and transportation is one of them. Of course, the airline is a major segment. I can talk a little bit about my background in the space.

We have people that have spent many years in the industry. And that greatly influences how we approach solutions. We know the pain points, the use cases, and can strategically influence business outcomes from the perspective of the client and industry. It’s about business impact and outcomes – we’re not marketing technology for technology’s sake.

As for my background, I come from the commercial side of an airline, United Airlines, where I led the business operations of MileagePlus, as well as United.com.

In terms of my philosophy, oftentimes technology is not the problem. It's more about envisioning from a business perspective of what opportunities to pursue. My experience as CIO at Amtrak, also gives a deep understanding of the technical challenges, is what I try to bring to our clients.

We also know that there's a convergence going on between business ideas and technology ideas. And oftentimes, the transformational discussion is coming from leveraging technology capabilities.

It's about new business capabilities being enabled by technology and innovation. It’s our job to help the understanding and also how it can be leveraged for a new purpose or value. To achieve success requires applying technologies, like AI, in a way that drives purposeful innovation. Whether it be customer service, profit, of course, revenue uplift, cost savings, we can adjust the focus to fit the specific need. This is the fundamental piece.

How does IBM perceive the opportunities and challenges for airlines seeking to integrate AI with their systems?

We understand that organizations are trying to figure out how to leverage AI. Frankly, they also are trying to figure out how to leverage their own internal data, and particularly in the airline sector. The fact is there is a lot of data in the airline industry, and most airlines have a major advantage in their proprietary data and decades of industry expertise to create significant change. But airlines have struggled to cross the data silos.

We see this as an opportunity for AI, but the big theme is around AI in helping organizations optimize processes, people, technology, and bring those together to truly rethink the business around those areas, and, ultimately, cultural change.

Is cognitive synonymous with machine learning?

You can easily get into the technical weeds, but to put it simply, there are multiple levels of AI.

Cognitive computing represents the multiple levels. It's basically bringing in a smarter version, or an intelligent version, of existing or new processes. Deep learning has brought learning capabilities into systems and machine learning elevates this by incorporating a level of reasoning, without explicit programming required.

The future continues to be very bright for AI and cognitive as we continue to move into broader and more general forms of AI, such as leveraging neuroscience. Organizations that seek to leverage AI and other exponential technologies are best positioned to capture transformational value and redefine competition and markets.

It’s also about the cloud. What we're seeing at IBM, with our hybrid multi-cloud strategy, particularly with our acquisition of Red Hat, is an opportunity to be that trusted partner for our clients.

For example, we helped American Airlines not just migrate critical applications to the cloud, but also helped transform their application development, organization and skills, all with the goal of responding better and faster to customer needs. We worked with the airline to create and roll out a dynamic rebooking app – a more convenient digital service for its customers – and empower employees with tools to change how they worked.

It’s a great example about how a technology, in this case, the cloud, wasn't just about the infrastructure; it was around empowering people, and accelerate features for AA.com, and even delivering enhancements to customers when airlines disruptions occur.

One of the technical challenges established carriers have is figuring out how to move from legacy tech systems to newer ones that can incorporate larger data sets and AI protocols. How do you help airlines with that digital transition?

One of the things we're really proud about, and this goes again back to cloud, is with the acquisition of Red Hat, we're the honest broker. We ensure companies are not locked into specific paths of cloud technology. While the cloud brings much capability, we've seen a heavy dependence on proprietary technology, and while we've been a long time open source provider and supporter, with Red Hat we’re re-focusing on open technology and helping clients avoid the risks of lock-in.

With Red Hat’s OpenShift container platform, we have elevated our conversation with the CIO and with IT organizations. OpenShift is a foundational component to transforming all of your business applications into hybrid and multi-cloud environments. This really accelerates the transformation if you have hundreds of applications, including many of which are based in older legacy technologies.

What is a “container platform?”

A container platform is all about building and scaling applications faster and more efficiently. Think about how the physical container in the transport industry fundamentally transformed standards and efficiency. It's similar from a technology perspective, that companies can build and modernize their applications into a container, and then be flexible in moving that container into the right cloud infrastructure, whether it be IBM Cloud, or other cloud providers. The value to CIOs, and we're working with CIOs aggressively, is to go fundamentally modernize and build your applications once, and take advantage of the container platform OpenShift to avoid cloud lock-in and deploy to the preferred cloud provider.

Another benefit of a container platform is ensuring strong security and consistency of the security approach. This is a critical aspect in any digital transformation, especially as we move into AI and hybrid multi-cloud technologies.  

We are proud to be a significant partner in the airline industry with specific innovations and solutions unleashing real value for passengers and companies. We have a tremendous history together, including still today we process roughly 80 percent of all reservation transactions using IBM technology.

In addition, we are one of the earliest strategic partners at IATA, going back to the early '80s. Perhaps you've heard the story about our partnership with American Airlines now Sabre on their initial reservation system. We've had multiple industry-level innovations, including the reservation system, kiosks for check-in, mobile suite of applications for employees and now our AI engine which we’ve introduced with Malaysia Airlines.  

We recognize the travel sector has about 10 percent of the world’s jobs, according to the World Travel & Tourism Council. So in our industry, and specifically at airlines, jobs and employees are critical. There are the onboard staff, including pilots and flight attendants, and ground staff, including operations and maintenance amongst others. These are highly skilled employees, but we haven't given them right the tools or access to other technologies like we've done with consumers to maximize their abilities.

Technology is driving new ways of working and fostering cultures of collaboration within the enterprise. In 2014, we began partnering with Apple, a collaboration based on an exclusive arrangement with IBM to build business applications for large enterprises. We've since created our airline portfolio of iOS apps that major global airlines across the world have deployed, including Singapore Airlines, United Airlines, KLM, and Air Canada.

Through our partnership with Apple, we’re changing how employees work by putting data and insights at their fingertips and enabling them to do their job better. We helped KLM, for example, make aircraft turnaround even more efficient with an app used on the apron that is a major contribution to helping an aircraft depart on time.

It's not just optimizing their individual work areas, but it’s about empowerment and the opportunity to collaborate across the silos, because oftentimes, a large organization like an airline does not necessarily have an environment that easily fosters collaboration. Roles tend to be walled off from each other, in a sense. We’ve found by facilitating empowerment and collaboration, employees will naturally provide better customer service at better costs as well as find ways to drive higher revenue. They're very motivated once you kind of give them the right tools.

What’s the process for developing these kinds of solutions for airlines?

We spend roughly about $6 billion a year in research. IBM Watson AI has been a key part of those efforts. Our research team has developed a very specific AI algorithm for the airline industry that essentially leverages appropriate customer and operational data to identify the propensity of spending and seat booking, and provides highly targeted offers resulting in higher uptake and a better passenger experience.  

With now 27 patents on this airline AI algorithm, now part of our IBM Travel Platform and called Personalized Pricing and Offers, we have garnered industry recognition, including receiving last year’s World Travel Award for the World’s Leading AI Travel Technology Provider. We are excited to extend our AI capabilities into additional use cases and with other innovative airline partners – especially as we seek to optimize other areas of the passenger journey, including managing disruptions.